基于形态学和小波变换的心电信号去噪算法

在处理心电信号采集过程中混入的基线漂移、工频干扰及肌电干扰等噪声的过程中,小波变换取得了广泛的应用。针对小波算法的缺陷及不足,提出了一种基于数学形态学和小波阈值的混合算法。该算法利用非线性形态学滤波器滤除基线漂移,将获得的含高频噪声心电信号通过小波阈值算法进行处理,最后获得无噪声的ECG（心电）信号。采用MIT/BIH Arrhythmia Database中的数据对算法进行了验证,实现了三种主要干扰的滤除,本算法效果良好,为后续特征点的识别奠定了基础。

An ECG Denoising Algorithm Based on Morphology and Wavelet Transform

In processing the ECG signal denoised by baseline drift,frequency interference and EMG interference,the wavelet transform had been widely used.Since the algorithm had some shorting in dealing with ECG signal,it proposed a mixed algorithm consisting of ECG wavelet threshold and morphology.This kind of algorithm got rid of baseline drift by nonlinear morphology,then the ECG signal containing high-frequency interference would be processed by wavelet threshold transform,in the end,it got the ECG signal with little noise.The MIT/BIH arrhythmia database was used to prove the algorithm,and the three main noises were removed effectively,the effect of the algorithm was good,and it lays the foundation of recognizing the ECG signal subsequently.